Project Outcomes Report

The main goal of our research was to advance our understanding of the behavior of complicated fault systems so we can better forecast future earthquake occurrence. Most current analyses of earthquake hazard rely on simple models that assume that fault slip and earthquake recurrence are relatively constant in time, and that the rate of future earthquake recurrence can be forecast by using average rates of energy storage and release along individual faults. Our research is motivated by the need to move beyond these simple models towards more physically realistic models that might allow us to more effectively forecast future earthquake occurrence. The most critical need in this regard is more comprehensive geological data sets that show how real faults have actually behaved in the past. To address this, we have studied all four major faults of the Marlborough fault system that comprise the Pacific-Australia plate boundary in northern South Island, New Zealand. Specifically, we documented the rate at which each of these faults release seismic energy in large earthquakes. These efforts have resulted in a large number of new fault slip rates, as well as new data on the ages and displacements in past earthquakes on all of these faults. Collectively, these new results allow us to “dissect” the fault slip and earthquake behavior of an entire system of mechanically inter-related faults that together accommodate most relative motion between the Pacific and Australia tectonic plates. The results demonstrate that many common assumptions used in current earthquake hazard models may not be correct, including the idea that fault slip and earthquake occurrence are relatively constant processes that occur at relatively regular rates.

Intellectual Merit

Our newly generated incremental fault slip rate and paleo-earthquake observations reveal  highly variable rates of earthquake occurrence on all four of the major faults we’ve studied, in marked contrast to the expected simple, monotonic behavior that underlies most current seismic hazard models. Thus, most basically, our results demonstrate that many common assumptions used in current earthquake hazard models may not be correct, including the idea that fault slip and earthquake occurrence are relatively constant processes that occur at relatively regular rates. Our results have added a large number of data to this previously data-limited discussion, providing useful data points that will facilitate comparisons with similar data collected from other fault system around the world. For example, our results have direct implications for furthering our understanding of how other fault systems store and release earthquake energy, perhaps most obviously the similar fault systems that underlie much of southern and northern California.

Broader Impacts

The three main goals of the Broader Impacts part of our study were to: (1) improve our understanding of the actual behavior of complex, plate-boundary fault systems, in the interests of improving next-generation models of probabilistic seismic hazard assessment; (2) educate and train graduate students at both USC and UCLA in state-of-the-art research techniques, preparing them for future careers in Earth Sciences research; and (3) strengthen our ongoing collaboration with scientists in New Zealand, thus ensuring the use of our data in improving seismic hazard assessment strategies in that country. All three of these efforts met with great success during our project. In terms of immediate impact, our ongoing collaborations with New Zealand government geologists have ensured that all of our data are now being incorporated in New Zealand’s National Seismic hazard model, strengthening seismic resilience in that country. This research formed the bulk of the thesis work of three USC graduate students, including two young female scientists. All three of these students have been trained in state of the art research techniques that have positioned them to succeed in their future research endeavors. In terms of advancing seismic hazard assessment, our new data provide basic information about how complex, multi-fault plate boundary fault system “trade off” in accommodating relative plate motion in time and space. In particular, our new data provide basic constraints over the time and displacement scales at which these interactions take place. These data, from a real fault system, thus constitute a base level of actual fault system behavior that can be used to constrain seismic hazard assessment models. An additional outcome of our research is the collection and public release of an extensive set of lidar data (highly detailed digital topographic data) that are now freely available to be used by other researchers and government agencies seeking to understand the major faults in northern South Island New Zealand specifically, and other, similar fault systems and geological processes around the world.

Last Modified: 01/30/2019
Modified by: James F Dolan